Aspiration failure - gov.uk€¦ · aspiration gap in rural Ethiopia, which goes hand in hand with...
Transcript of Aspiration failure - gov.uk€¦ · aspiration gap in rural Ethiopia, which goes hand in hand with...
Aspiration failureA poverty trap for indigenous children in Peru?
Laure Pasquier-Doumer and Fiorella Risso Brandon
Wo
rkin
g P
aper
DECEMBER 2013
116
www.younglives.org.uk
Wo
rkin
g P
aper
DECEMBER 2013
116
www.younglives.org.uk
Aspiration failureA poverty trap for indigenous children in Peru?
Laure Pasquier-Doumer and Fiorella Risso Brandon
This paper was presented at a conference on Inequalities in Children’s Outcomes in Developing Countries hosted by Young Lives at St Anne’s College, Oxford on 8-9 July 2013.
http://www.younglives.org.uk/news/news/children-inequalities-younglives-conference-2013
The data used come from Young Lives, a longitudinal study of childhood poverty that is tracking the lives of 12,000 children in Ethiopia, India (in Andhra Pradesh), Peru and Vietnam over a 15-year period. www.younglives.org.uk
Young Lives is funded from 2001 to 2017 by UK aid from the Department of International Development and co-funded by the Netherlands Ministry of Foreign Affairs from 2010 to 2014.
The views expressed are those of the author(s). They are not necessarily those of the Young Lives project, the University of Oxford, DFID or other funders.
Aspiration failure: A poverty trap for indigenous children in Peru?
Laure Pasquier-Doumer and Fiorella Risso Brandon
First published by Young Lives in December 2013
© Young Lives 2013 ISBN: 978-1-909403-29-1
A catalogue record for this publication is available from the British Library. All rights reserved. Reproduction, copy, transmission, or translation of any part of this publication may be made only under the following conditions:
• withthepriorpermissionof thepublisher;or
• withalicencefromtheCopyrightLicensingAgencyLtd., 90TottenhamCourtRoad,LondonW1P9HE,UK,orfromanothernationallicensingagency;or
• underthetermssetoutbelow.
This publication is copyright, but may be reproduced by any method without fee for teaching or non-profit purposes, but not for resale. Formal permission is required for all such uses, but normally will be granted immediately. For copying in any other circumstances, or for re-use in other publications, or for translation or adaptation, prior written permission must be obtained from the publisher and a fee may be payable.
PrintedonFSC-certifiedpaperfromtraceableandsustainablesources.
Young Lives, Oxford Department of International Development (ODID), University of Oxford,
Queen Elizabeth House, 3 Mansfield Road, Oxford OX1 3TB, UK
Tel: +44 (0)1865 281751 • E-mail: [email protected]
Funded by
1
Contents
Abstract 2
Acknowledgements 2
The Authors 2
1. Introduction 3
2. Data, concepts and definitions 8
The data 8
How do we define indigenous people? 9
How do we define aspirations of the children? 10
How do we classify aspirations? 11
How do we measure aspiration gap? 12
3. Which occupations do Peruvian children aspire to? 13
4. How are aspirations shaped? 15
Empirical strategy 15
Results 17
5. The effect of the aspiration gap on educational attainment 19
6. Conclusion 22
References 24
Appendix 28
2
Abstract
This paper aims to contribute to understand the mechanisms underlying the complex
exclusion process of indigenous people in Peru, by analysing the role played by aspirations in
the investment in education of indigenous children. To address these issues, the paper relies
on a very rich data set, the Young Lives data, and use an original instrument that allow to cast
light on the causal relation between aspiration and educational outcomes. We find that
aspiration failure is a channel of inequality persistence between indigenous and non-
indigenous people, but that aspiration failure do not takes the form of a lack of aspiration.
Indigenous children do not have internalized racial schemas about occupation or about their
opportunities. However, the gap between their aspiration and their current socio-economic
status is too large, in so far as it has a disincentive effect on forward-looking behaviour.
Acknowledgements
The authors thank Javier Herrera and Javier Escobal for fruitful suggestions and stimulating
discussions, and for their help in the access to the data. We also thank participants in the
Young Lives Conference “Inequalities in Children’s Outcomes in Developing Countries”
(Oxford, July 2013) for their comments. This paper has received funding from the NOPOOR
project (www.nopoor.eu) under the FP7 of the European Commission.
The Authors
Laure Pasquier-Doumer holds a PhD in development economics and is researcher at DIAL,
a research unit of the French Research Institute for Development (IRD) and the University of
Paris Dauphine. Her research focuses on inequality of opportunity in the labour market and in
education.
Fiorella Risso Brandon is a PhD candidate in Economics at University Paris Dauphine. Her
areas of expertise include poverty, inequality of opportunity, and discrimination in Peru.
3
1. Introduction
Since the last decade, Peru has experienced dramatic poverty alleviation, in both rural and
urban areas and for various dimensions of poverty. Although incidence of poverty among
indigenous people also decreased, it has not reduced in relative terms. The gap between
indigenous and non-indigenous people remains as high as it was 10 years ago. Hence, the
monetary poverty rate among indigenous people is 1.6 times higher than among non-
indigenous people in 2010, reaching 42.5% compared to 25.8% for non-indigenous people,
while it was 1.3 times higher in 2004 (INEI 2011).
Intergenerational poverty trap for indigenous people is widely acknowledged and documented
in Peru (Escobal and Ponce 2007; Cueto et al. 2009; Trivelli 2005; Figueroa 2006; Barrón
2005; Ñopo, Saavedra and Torero 2004; Mac Isaac 1994). These studies lead to the conclusion
that exclusion, understood as discrimination in access, is the main driver of inequality. In
addition, exclusion mechanisms operate mostly when individuals are acquiring skills. While
discrimination in the labour market seems to be low, coming from an indigenous family is an
important predictor of low educational outcomes, in terms of repetition, dropout or attainment
(Cueto 2007, Escobal and Ponce 2007). Barrón (2008) estimates that partial reductions of
exclusion in the access to education would reduce inequality as much as the complete
elimination of discrimination. Pasquier-Doumer (2002) stress the structural dimension of the
inequality in access to education between indigenous and non-indigenous people, as the gap
in access to secondary or high education between the two groups has widened throughout the
second half of the twentieth century.
However, very little is known on how the exclusion mechanisms of indigenous people
operate. Cueto et al. (2009) show that poverty status, geographical localization, or other
characteristics of indigenous households only explain a marginal portion of the low
educational achievement of indigenous children, while educational process and peer effects
play a crucial role. But many unanswered questions remain on how the educational process
and the clustering effect lead to the exclusion of indigenous children.
This paper aims to contribute to understand the mechanisms underlying the complex
exclusion process of indigenous people in Peru, by analysing the role played by aspirations. In
a common definition, aspiration is a desire or ambition to achieve something. This concept
suggests that some effort would be exerted to realize the desired aim or target. Following
4
Bernard and Tafesse (2012), “aspirations combine or summarize the preferences maintained,
the beliefs held, and possibly the constraints acknowledged by an individual about aspects of
the future” (p.3).
Studying aspiration as a channel of intergenerational transmission of inequalities results of a
long tradition in psychological and sociological literature. 1
In particular, a strand of research
focuses on the shaping of aspirations for ethnic minorities and analyses what is the effect of
aspiration of ethnic minorities on their outcomes (Kuvlesky and Patella 1971, Cosby and
Picou 1973, Kao and Tienda 1998; Qian and Blair 1999; Goldenberg et al. 2001; St-Hilaire
2002; Bigler, Averhart and Liben 2003, Glick and White 2004, Zhang et al. 2011). Most of
these studies conclude that ethnic belonging is a strong predictor of aspirations, which in turn
influence outcomes. Results are more controversial when socio-economic status is properly
taken into account. However, all of these studies are conducted in the USA and related to
Black-, Hispanic- or Asian-American minorities.
More recently, a growing economic literature addresses the relationship between poverty,
inequality and aspirations. This strand of research, inspired by behavioural economics, aims to
overcome the neo-classical analysis of poverty in which decision-making is only shaped by
structural constraints that poverty imposes. As stressed by Duflo (2006) “what is needed is a
theory of how poverty influences decision-making, not only by affecting the constraints, but by
changing the decision-making process itself” (p.378). Recent theoretical developments and
some evidences support the idea that aspirations is of principal concern for poverty reduction
and equality of opportunity.
While some authors as Piketty (1998) introduced prospect for future social status in the
formalization of the dynamic of inequalities, the concept of aspiration as such is introduced
for the first time in the economic literature by Ray (2006). Inspired by the work of the
anthropologist Appadurai (2004) on ‘the capacity to aspire’, Ray defends the idea that poverty
and the failure of aspirations are reciprocally linked in a self-sustaining trap. For these
authors, aspirations are socially determined. Ray defines the concept of ‘aspirations window’
that one forms from lives or achievement of ‘similar’ or ‘attainable’ individuals. The effect of
aspiration on behaviour is then channelling by the ‘aspiration gap’, defined as the difference
1 In line with Marxist sociology, Reissman (1953) intended in a pioneer work to formalize the link between aspiration and
social class. In the 1970’s, various social psychology studies are conducted to understand how aspiration affect child
development and attainment.
5
between the standard of living that is aspired to and the standard of living that one already
have. In case of very small or very wide aspiration gap, individuals have little incentive to
raise standards, because the distance to fill the gap is too small or too large. Ray adds that for
a polarized society such as that of Peru, the aspiration gap of poor people will be too large if
the poor do aspire to be like the rich or too small if they do not include the rich in their
aspiration window.
Along similar line, Page (2005) offers a formalization of the causal link between aspirations
and educational behaviour using the notion of reference point from prospect theory. In a
model of human capital accumulation, Mookherjee, Napel and Ray (2010) emphasize
aspirations formation as a key factor of accumulation incentives. By introducing local
complementarities in the aspirations formation -social interactions with skilled neighbours
raise parental aspirations for their children-, they are able to explain inequality persistence
without hypothesis of capital-market imperfection or spatial mobility of agents. As argued by
the authors, social networks based on ethnicity can play a role analogous to those of one’s
physical neighbours in their model.
Beside theoretical developments, recent evidences related to developing countries suggest that
integrating aspiration formation in the understanding of poverty trap and inequalities
persistence is of principal concern. Bernard, Dercon and Tafesse (2011) observe a small
aspiration gap in rural Ethiopia, which goes hand in hand with low self-efficacy. Aspiration
gap is measured as one’s belief that she could become as successful as her role model within
five years. In addition, aspiration failure and low self-efficacy correlate with underinvestment
as they lower demand for long-term credit and the productive use of this credit. Relying on
qualitative and quantitative data, Camfield et al. (2012) find evidence of an adaptation process
in Thailand, which leads the poor to stop aspiring to what they cannot achieve. As argued by
the authors, this aspiration failure could explain why objective poverty does not affect
people’s satisfaction with their lives in Thailand. With qualitative interview in Egypt, Ibrahim
(2011) highlights intergenerational transmission of aspirations’ failure for poor people, which
could be a channel of inequality persistence. In Peru, Dercon and Krishnan (2009) explores
the relationship between material poverty and educational aspiration of children using the
Young Lives data. They find a strong and positive correlation between material circumstances
and educational aspirations.
6
However, these studies fail to address two main questions. First, what are the mechanisms by
which aspirations are shaped in developing countries? Are aspirations culturally or socially
determined as argued by Ray (2006), Appadurai (2004), or as suggested by Sen (2004), who
defends the idea that culture plays a major role in value formation? Or are aspiration shaped
by external constraints through an individual adaptation process? In other words, do
aspirations differences between two groups result from cultural differences between the two
groups or do they only reflect different level of constraint? The first hypothesis will be called
the ‘internal channel hypothesis’ later on, and the second one the ‘external channel
hypothesis’. To clearly distinguish these two channels, let’s quote Rao and Walton (2004):
“an economist may find it helpful to think of this internalization of the perceived possibility of
success or failure as a constraining preference that interacts with exogenous constraints to
affect human action (…). The incest taboo, for example, is a feature of most human societies.
Most people would not consider breaking it, not just because of fear of social sanctions, but
simply because the taboo is so deeply ingrained within their psyches. Thus the taboo is
simultaneously an inherent preference against incest and a social constraint.” (p.15).
Getting back to our main concern, how ethnic group can shape aspirations? Following the
‘internal channel hypothesis’, being indigenous may lead to an aspirations failure if
indigenous people have internalized the discriminatory values of the ‘criolla’2 elite, and then
their objective chances of attaining a “high” socio-economic status. This idea is supported by
a recent literature that highlights the link between current inequality and the colonial past of
Latin American countries through a ‘path dependence’ process (Acemoglu, Johnson and
Robinson 2002, Mahoney 2003, Engerman and Sokoloff 2005). Internalization of
discriminatory value is also observed by Hoff and Pandey (2006) in India. Relying on an
experiment, they show that making caste salient hurt low caste performance. In Peru, the
symbolic prestige of being ‘white’ is widely acknowledge as a legacy of the colonial era
(1514-1821). Some evidences show that it generates racism and socio-racial discrimination,
and that indigenous people have internalized these discriminatory values (Portocarerro 1993,
Henríquez 1995, Ames 2012). Thus, this process defined by Bourdieu (1997) as "symbolic
violence" may limit the aspiration window of indigenous people and perpetuate inequalities.
On the other hand, being indigenous is also associated with other characteristics, such as
being poor or living in a rural environment. These characteristics or ‘external constraints’ may
2 ‘Criolla’ or ‘criollo’ refers to people who is Spanish descent but is born in Spanish-America.
7
be the main determinant of an aspiration failure, as they limit access to information or
opportunity to invest in the future. As instance, following the ‘external constraints’
hypothesis, the children may not aspire to be a doctor because they know that their parents
would not have the funds to pay their studies. With the ‘internal channel hypothesis’, they will
not aspire to be a doctor because they think that a doctor has to be ‘white’ or that they are not
smart enough to succeed at medical school. In the line of the ‘external channel hypothesis’,
some authors highlight major transformations of the Peruvian which allowed some indigenous
people to relax their constraints and to consequently benefit from economic success (De Soto
1986, Golte and Adams 1987, Adams and Valdivia 1991, Hubert 1997, Mendoza 1997).
These transformations are the integration of indigenous people to the democratic process3, the
huge development of education, and the massive migration to urban areas.
The policy implications of the ‘internal channel’ and ‘external channel’ hypothesis are
dramatically different. In the latter case, ‘leveling the playing field’ for indigenous people –to
quote the metaphor that Roemer (1998) used to define equality of opportunity- would be
enough to reduce the persistence of inequalities. In other words, proving equal access to
human and physical capital to indigenous people, by alleviating the exogenous constraints
they face, would be efficient. In the former case however, these policies will not be sufficient
to break the vicious circle of poverty for indigenous people.
Second, another gap in the existing literature is that the causal effect of aspiration on forward-
looking behaviour is not well established. Most of the studies linking aspiration and poverty
rely on cross-sectional data, and consequently are not able to address the endogeneity issue
due to reverse causality and unobserved variables. Indeed, unobserved characteristics may
jointly determine aspiration and forward-looking behaviour, and aspiration may be adjusted to
outcomes of looking-forward behaviour. Studying the causal impact of aspiration on
educational choices, Page, Levy Garboua and Montmarquette (2007) overcome the
endogeneity problem by using an experiment. However, their experiment may seem very far
from the purpose they aim to analyze. It consists of asking individuals to solve a given
number of anagrams. Individuals are reward if they reach different levels of difficulty. Indeed,
in this experiment, time and money investments to perform the task are proxy of educational
choices and the expected gains of the game relative to an initial sum reflect the aspiration gap.
In addition, their experiment is conducted in France. The link between aspiration and
3 The 1979 Constitution provided for universal suffrage and allowed illiterate people to vote, knowing that the
bulk of indigenous people was illiterate.
8
behaviour may be different in Peru than in France, because of a high constraint level for the
poor due to market imperfection, and because the social stratification is partly shaped by the
colonial past of Peru.
Based on these perspectives, this paper investigates first whether aspirations of indigenous
people differ from the ones of non-indigenous people in Peru, and if so, through which
mechanisms. It seeks to identify the respective relevance of the ‘internal channel hypothesis’
and the ‘external channel hypothesis’ in the Peruvian context, using a dynamic approach of
aspiration formation. Then, the paper aims to cast light on the causal relation between
aspiration and educational success. It encompasses the question whether aspiration failure
leads to underinvestment in education. To tackle these issues, the paper relies on a very rich
data set, the Young Lives data.
Section 2 presents the data, and defines the concepts used in the paper. Section 3 describes the
aspirations and the aspiration gap of the children, and their dynamic. Section 4 examines the
determinants of aspirations and the role of ethnic group. Section 5 analyses the causal link
between aspiration failure and educational outcomes. Section 6 summarizes and concludes.
2. Data, concepts and definitions
The data
We used survey data of the Young Lives project. Young Lives (YL) is a longitudinal study of
childhood poverty for understanding its causes and consequences, and in order to influence
the formulation of public policies. It is led by a team in the Department of International
Development at the University of Oxford. In association with other partners, the data are
collected in four countries: Ethiopia, India (Andhra Pradesh), Peru4 and Vietnam. It follows
over 15 years two cohorts of children, the Younger cohort and the Older cohort, born in 2001-
02 and 1994-95 respectively. They were first surveyed in 2002, and then every four years. So
far, the data are available for three rounds: 2002 (R1), 2006 (R2) and 2009 (R3).
4 In Peru, the project is implemented by three institutions: Instituto de Investigacion Nutricional (INN), Grupo de
Analisis para el desarrollo (GRADE) and Save the Children UK.
9
The YL data are unique among developing countries. Indeed, through interviews with large
cohorts of children and their caregiver, a wealth of information are collected in a dynamic
way, not only about the material and social circumstances of the children, but also about
perspectives on their lives and aspirations for the future, as on the environmental and social
realities of their communities. In particular, for the older cohort, children are interviewed on
their educational and occupational aspiration at each round.
In Peru, the two cohorts of children were selected using a multi-stage, cluster-stratified and
random sampling, and with a pro-poor bias, ensuring that the sample represents 95% of the
population excluding children in the richest 5%. YL households are found to be very similar
to the average household in Peru, although they may have better access to some services
(Escobal and Flores 2008). In this paper, we use the post-stratification proposed by Escobal
and Flores (2008), to minimize the potential bias.
In this paper, we focus on the older cohort, surveyed at 8, 12 and 15 years old. This cohort
includes 714 children in 2002, 685 children in 2006, and 678 children in 2009. The low
attrition rates between the three rounds (5% between 2002 and 2009) are a key factor of
quality of these data as well.
How do we define indigenous people?
Ethnic group is a complex and controversial concept. It is multidimensional, as an ethnic
group is a group of people whose members share some common characteristics which could
be both cultural, as language or religion, and geographical as a place of origin or a territory. It
is a moving concept in time and space also. Therefore, it is problematic to define ethnic
belonging with only one criterion. In Peru, the term indigenous commonly refers to the people
belonging to Andean ethnic groups Quechua and Aymara, and the Amazonian ethnic groups.
However, the huge transformations experienced by the Peruvian society, as mestizaje and
migration, have made this concept quite fuzzy, as stressed by many authors.
In the YL data, ethnic group can be defined by self-identification or by ascription. Self-
identification measure does not reduce the definition of ethnic group to one dimension. But it
is widely acknowledged that self-identification underestimates the size of indigenous
population in the Peruvian context (Ñopo, Saavedra, and Torero 2004), where discriminatory
10
values against indigenous people are internalized, as already stressed. However, in the YL
data, it has been asked to the caregiver to define her ethnic group and the one of her child, by
choosing one answer from the following list: “White, Mestizo (include Andean indian),
Native of Amazon, Negro, Asiatic, other”. Following this definition, 97% of the population is
indigenous. The unclear boundaries of the Mestizo concept and maybe the willing to avoid
this sensitive question may explain that the bulk of interviewed people defines herself as
mestizo. In any case, this variable seems not reliable to define indigenous people here.
The language is the criterion most widely used in the empirical literature for identifying
indigenous people. Beyond the fact that it is easy to define and measure, language is a central
element of a culture and process of social cohesion. According to the United Nations,
language, and especially the mother tongue, is a key variable for identifying ethnic groups.
More than 40 indigenous languages are spoken in Peru, most in the Amazon and with a
relatively small number of speakers. Quechua is the most frequent indigenous language and is
spoken widely in the Andes. Aymara is the second most frequent indigenous language,
spoken mostly in the Southern part of the Peruvian Andes. However, the language criterion
not only reduces ethnic belonging to one dimension, it also underestimates the indigenous
population as shown by Ñopo, Saavedra, and Torero (2004), mostly because of migration to
urban areas that reinforce learning Spanish. To mitigate these limitations, we widen the
definition to the grandparents tongue. Children are defined as indigenous if the first language
of one of their parents (mother, father or caregiver) learned as a child is Quechua, Aymara or
a language of the Amazon. Following this definition, 48% of the children are indigenous in
our sample.
How do we define aspirations of the children?
The concept of aspirations is multidimensional. As recalled by Bernard and Tafesse (2012),
aspirations span multiple and potentially interrelated dimensions. Individuals may have
wealth or income aspirations, educational aspirations, social status aspirations, or aspirations
about others such as their children. However, the vast majority of studies that address the
issue of aspirations restrict the definition of aspirations to one dimension. The reason for this
is mainly the difficulty to aggregate various dimensions of aspirations into a single indicator.
The dimensions most frequently tackled are income aspirations and educational aspirations.
Both have the advantage that they can be measured by an ordered indicator.
11
In this paper, we prefer to focus on occupational aspirations of the children for two main
reasons. First, it is widely acknowledged that the division of labour is a key factor of social
inequality. Occupation reflects very closely the social status, which integrates the notion of
wealth, but also of power and prestige in its Weberian conception. Occupation is therefore a
more multi-dimensional concept than income or education. The second reason for studying
occupational aspirations is that educational aspirations are very high in the context of Peru,
whatever the characteristics of the families. This phenomenon is well known in Peru as the
“education myth”. Ansión et al. (1998) highlight that in Lima, 95% of parents want that their
children attain a grade higher than secondary. In the YL survey, around 90% of the 12 aged
children want to reach a post-secondary level and 77% want to complete university. Because
of this lack of variance, it seems not appropriate to use educational aspirations in the Peruvian
context to understand the determinants of aspirations.
In the first two rounds of the YL’s survey, children have been asked what they want to be
when they grow up. In the R1, children had to choose an occupation from a list of seven
modalities but could also add another one. Because a quarter of children listed an occupation
out of the list, a total of 40 different occupations have been mentioned. In the R2, the list was
larger, including 32 modalities. At the end of the day, 43 occupations have been proposed in
the R2. Unfortunately, the wording of the question has been changed in the R3, to capture
children expectation rather than children aspiration. For this reason, the R3 is not used in this
paper to measure aspiration. It will be only used to measure educational outcomes.
How do we classify aspirations?
The difficulty of conducting quantitative analysis based on occupation is to classify them in
an ordered way. To face this difficulty, we use a measure of socio-economic status that is
commonly used in sociology. Indeed, recognizing that occupation is the main dimension of
social stratification, stratification researchers have developed ways to derive status measures
from information on occupations. We adopt the same approach as Ganzeboom and Treiman
(1996) to calculate the International Socio-Economic Index of occupational status. In this
approach, scores are created by computing a weighted sum of socioeconomic characteristics
related to each occupation, usually education and income. It has been widely acknowledged
12
by sociologists that this approach allows capturing the basic parameters of the process of
stratification.
First, occupational aspirations have been recoded using the 3-digit International Standard
Classification of Occupation 2008 (ISCO-08). In a second step, we assign a score to each
occupation, which is a linear combination of the average level of education required by the
occupation and the average income it generates. To compute this score, we use the Encuesta
Nacional de Hogares (ENAHO), a representative survey of the Peruvian population carried
out in 2006, with around 62.000 individuals. The average level of income and the average
years of education of one occupation have been standardized and added up with the same
weight, as the various procedures used in the literature to derive the weights conclude all to
more or less the same weight for education and income (Ganzeboom and Treiman 1996).
The scores of the occupation that children aspire to are used as a metric measure of the social
status the children aspire to, which is ordered according to education and income criteria.
Let’s denote it as . As a robustness check and to take into account non-linear effect, we
also use an ordered-categorical measure of aspiration, denoted , which is calculated by
aggregating occupations into four groups, defined as the quartiles of the score from all the
occupations listed by the children in R1, and R2. takes the values 1 to 4. Table 1
presents the average income, the level of education and the standardized score for each group
and Table 2 gives the composition of each group in terms of occupations. The most
representative occupation for the low aspirations group ( ) is “agricultural worker”,
while it is “driver” or “mechanic” for the intermediate aspirations group ( ), “teacher”
for the high aspirations group ( ) and “medical doctor” or “engineer” for the very high
aspirations group ( ).
[Tables 1 and 2 here]
How do we measure aspiration gap?
Following Ray (2006) aspiration gap is defined as the difference between the social status that
is aspired to and the social status that one already have. As the social status that the children
aspire to is measured by the score of occupational aspirations , we would like to use the
occupation of the parents to measure the actual social status of the children in a comparable
13
way. Unfortunately, only sectors of activity of the parents are collected in the YL data. That
the reason why we choose to use the education of the parents and the consumption of the
household to approximate the education required and the income provided by the occupation
of the parents. More precisely, the score of actual social status of the household, denoted as
, is a linear combination of the level of education of the most educated parent and the
consumption per capita of the household. As for the score of aspiration, these two variables
have been standardized and added up with the same weight. The aspiration gap is denoted as
, and is then defined as:
(1)
3. Which occupations do Peruvian children aspire to?
At 8 years old, the most desired occupation is teacher, as 31% of the children want to become
teacher (Table 3). Medical doctor ranks second. This result is consistent with the results of
Benavides et al. (2006), whose find that Peruvian children mainly aspire to become teacher
and doctor in rural area. Frequent interactions with doctors and teachers and the respect
inspired by these professions may explain this result, as found by Oluigbo (1976) in the
context of Nigeria. Fire-fighter and policeman are desirable occupation as well and ranks
third, ahead of engineer and farmer. At 12 years old, teacher is still the most popular
occupation, ahead of engineer, doctor, nurse and lawyer. Farmer is among popular occupation
at age 8 but becomes unpopular as children grow up (7% of the children 8 years want to
become farmer, compared to less than 3% when they are 12 old). Medical doctor and engineer
are among the most desirable occupations for all ages. The desire to become a doctor is rather
stable with age, although decreasing slightly, as the one to become nurse or fire-fighter or
policeman. Engineer, accountant and lawyer become more attractive as children grow up.
Finally, the number of listed occupations increases with the age of the children. Taken
together, these results suggest that the aspiration window of the children get wider as they
grow up. They enrich their knowledge on the labour market, on the living standards provided
by these occupations, but they may also change their criteria to evaluate an occupation.
[Table 3 here]
14
Another interesting result from Table 3 is the discrepancy between genders as far as the
choice of the desired occupation is concerned. Girls 8 years aspire to become nurse when
boys want to be fireman or policeman. At 12 years old, secretary is among the top five
favorite occupations for girls, while boys prefer to become a mechanic or a lawyer. In
addition, girls tend to aspire to a narrower range of occupations than boys. This evidence is
well documented in the psychologist literature, where it is argued that girls believe many jobs
are unsuitable for girls, whereas boys have a greater occupational understanding and see more
occupational opportunities than girls (Reid and Stephens, 1985). However, this difference is
reduced if we take into account the lower number of girls in the sample, and if we suppose
that the number of listed occupations must be proportionate to the number of children
answering to this question.
Moving to the emphasis of this paper, comparison of aspirations of indigenous and non-
indigenous children shows that the top five favorite occupations of both groups are almost the
same at 8 and 12 years old (Table 3). However, the ranking differs between the two groups.
At 8 years old, the most desired occupation is teacher for indigenous children (41%) while
non-indigenous children most often aspire to be a doctor (31%). These preferences stand for
both sexes. At 12 years old, teacher is the preferred occupation for both groups, but the
proportion of children declaring this occupation is nearly twice higher among indigenous
children (31 against 18%). Another result is that the number of occupations mentioned is
smaller among indigenous children. But the difference is not significant anymore when we
relate this number to the proportion of indigenous children in the sample.
By ordering the occupations mentioned as aspiration, and then by aggregating them in four
groups of aspirations, as defined above, the differences between indigenous and non-
indigenous children remains rather small (Table 4). First, the average aspiration score of
indigenous children does not significantly differ from the one of non-indigenous children,
although it is slightly lower at age 8. Second, indigenous children do not show more often low
or intermediate level of aspirations. On the other hand, the share of children whose aspire to a
“very high” occupation is significantly lower for indigenous children at ages 8 and 12, and
indigenous children aspire significantly more often to a “high” occupation in contrast.
[Table 4 here]
15
The few differences between the two groups of children at two points in time may hide large
differences in the changes of the levels of aspirations as children grow. We therefore examine
the level of aspiration in a dynamic way. Table 5 compares the rate of upward mobility,
downward mobility and immobility between age 8 and 12 for indigenous and non-indigenous
children. Here again, the differences between indigenous and non-indigenous children are
small. Indigenous children lower less often their aspirations than non-indigenous children as
they are growing up, and revise their aspirations upwards more frequently, but the differences
are not significant.
[Table 5 here]
In contrast, indigenous children differ tremendously according to their aspiration gap (Table
4). The aspiration gap is nearly two times higher for indigenous children than for non-
indigenous children. This is mainly due to the lower current social status of indigenous
children. Following Ray (2006), the very large distance to fill the gap may impede indigenous
children to adopt forward-looking behaviours. But at this stage, it is difficult to have a
normative view on what it is a “too wide” aspiration gap.
4. How are aspirations shaped?
We now examine the channels by which aspirations are shaped, focusing on the role of ethnic
group in this process. In particular, we aim at testing the ‘internal channel hypothesis’ which
stipulates that ethnic belonging determines specific behaviour as a result of a ‘path
dependence’ process. Following this hypothesis, we expect that being indigenous is a
significant determinant of the aspiration shaping, once external constraints are taking into
account, because of an internalization of the discriminatory values conveyed by the elite since
the colonial past.
Empirical strategy
The first step of our empirical strategy involves estimating the level of aspiration by
introducing the ethnic group and proxies of external constraints as explanatory variables,
following the equation (2):
(2)
16
where
is the level of aspiration of the child i at the time t measured by the
score of aspiration, but also by the group of aspiration to take into account potential non-
linear relationship between aspiration and ethnic group; is a dummy that takes the value
one if the child is indigenous; is a vector of variables which approach the external
constraints of the child; is a vector of control variables; is the residual term and
the coefficients.
If indigenous children incorporate lower opportunities in shaping their aspiration due to racial
discrimination or internalize discriminatory values concerning their ability to reach some
occupations, we expect that the coefficient of being indigenous is significant and negative.
includes economics and informational constraints. Economic constraint is measured by a
wealth index5. Information constraint is measured by a dummy taking the value 1 if the child
lives in a rural area, as we suppose that access to information is more limited in rural than in
urban areas. Education of the parents captures both economic and information constraint, as it
is highly correlated with the occupation of the parents and it determines the cultural
environment of the child. Education of the parents is measured by the number of years of
education of the most educated parent. We also introduce the sex of the child and the sex of
the household head as control, because many studies show that aspirations are gender oriented
(McMahon and Patton 1997).
Models 1 to 4 estimate the score of aspiration by using an OLS specification at age 8 (models
1 and 2), and at age 12 (models 3 and 4), first without control variables (models 1 and 3), and
then with all control variables (models 2 and 4). Models 6 to 9 estimate the probability of
aspiring to one of the four occupational groups, by using ordered probit models. This
probability is estimated at age 8 (models 6 and 7), and at age 12 (models 8 and 9), with and
without controls (models 7 and 9, models 6 and 8 respectively).
Secondly, following the psychologist literature, we consider that the development of
occupational orientation and aspiration is a dynamic process. It occurs over the life course,
through interaction with the social environment. It is based on growing knowledge of self and
of occupations, and the reasoning about the relationship between the two that occurs with
cognitive development. Gottfredson (1981) defines four stages in the development of
5 For more details on the wealth index, refer to Escobal et al. 2003.
17
occupational aspirations. Around ages 9-13 years, children enter the third stage, called the
stage of orientation to social valuation, “when the more abstract self-concepts of social class
and ability become important determinants of social behaviour and expectations” (p.549).
Children begin to recognize prestige differences among jobs as well as social class. Some
sociological studies confirm that differences by social origin in shaping aspiration grow with
age, and that the awareness of social class goes hand in hand with the awareness of the ethnic
discrimination (Jacobs et al. 1991). Aspirations tend to decline as children grow and
understand the constraints imposed by their choices and the environment, such as ethnic
barriers. Consequently, they adjust their aspirations to feasible possibilities. This phenomenon
is also known as adaptive preferences in the economic literature (Elster 1985). We then
estimate the equation (3):
(3)
Where measures downwards adaptation of aspirations with {
If the ‘internal channel hypothesis’ is true, we expect that indigenous children lower more
their aspiration between ages 8 and 12 than non-indigenous children, and then that is
positive and significant (models 5 and 10). Table 6 presents the results of the estimations of
models 1 to 10.
Results
At age 8, being indigenous has a significant and negative effect on the level of aspiration, but
this effect does not hold when the socio-economic background of the children is introduced
(models 1, 2, 6, and 7). At age 12, we find no significant effect of the ethnic group on the
level of aspiration, except in the non-linear specification (models 3, 4, and 8). But this effect
disappears once controls are introduced (model 9). Hence, being indigenous is not a
significant factor of the aspiration level, once the level of external constraints is controlled
for. In addition, we do not find that indigenous children lower more their aspiration between
ages 8 and 12 than non-indigenous children. On the contrary, we find a negative effect of
being indigenous on the probability to lower aspiration between age 8 and 12. This effect is
significant as far as the aspiration level is measured by the group of aspiration (model 10).
Indigenous children do not are more aware of racial discrimination when they grow up.
Compared to non-indigenous children with the same external constraints, they tend to revise
upward their aspiration more often.
18
Economic and education background has a significant effect on the aspirations level on the
other hand (models 2, 4, 7, and 9). The wealth of the household has the main effect on
shaping the aspiration. This effect is positive, as the one of the education of the parents.
Children who have less educated parents or who are poor lower their aspiration, whatever
their age. They may exclude some prestigious occupations because they think that these
occupations are out of their reach, given their actual status, or because they do not know that
such occupations exist. When growing up, the role of wealth becomes more important in
shaping aspiration. The children may be more conscious of the financial resources needed to
pursue jobs that require high level of education while getting older. This adaptive process to
economic constraint is confirmed by the negative and significant effect of the wealth index in
the estimation of the probability of lowering her aspiration between age 8 and age 12 (models
5 and 10).
[Table 6 here]
Living in a rural area has a negative effect on the level of aspiration as well, in particular at
age 8 (models 2, 4, 7, and 9). This result can be interpreted as a limited access to information
on occupations in rural areas, as fewer occupations are represented there, in particular in
remote areas. By growing up, children have easier access to media and consequently increase
their knowledge on occupation and partially enlarge their aspiration window. Unlike to other
previous research related to developed countries, we do not find that girls restrict their
aspiration when compared to boys because they believe many jobs are not adequate for girls
(McMahon and Patton 1997). We find, on the contrary, that girls have higher level of
aspiration at age 8, other things being equal (models 2 and 7), but this effect disappear when
girls grow up (models 4 and 9), as girls lower more their aspiration between age 8 and 12 than
boys (model 5).
Taken together, these results call for the rejection of the ‘internal channel hypothesis’ which
would lead to a lack of aspiration for indigenous children. It appears that indigenous children
do not have internalized racial schemas that incorporate beliefs about occupation or about
their opportunities.
19
5. The effect of the aspiration gap on educational attainment
Following Ray (2006), individuals have little incentive to raise standards in case of very small
or very large aspiration gap, because the distance to fill the gap is too small or too large. The
question we address in this section is then whether the aspiration gap for indigenous children
is too large, in so far as indigenous children becomes discouraged from providing effort at
school. It may append if the effort needed to reach the education level that is required for the
occupation they aspire to are too costly. “The overall journey is too long, and therefore not
worth undertaking in the first place” (Ray 2006, p. 412). We then estimate the causal effect of
the aspiration gap at age 8 on the school performance between age 8 and 15. Following the
psychologist literature, we expect that the aspiration gap could have a different effect on the
educational performance while children grow up, as children are more aware of the education
level required for the occupation they aspire to. For this reason, we also estimate the effect of
the aspiration gap at age 12 on the school performance between age 12 and 15.
Children’s educational performance is measured by the number of repetitions between age 8
and 15, and between age 12 and 15. We choose this variable rather than the score in language
or in mathematics at age 12 and age 15, because it better reflects the evaluation by the school
of the progress of the child, and consequently is more linked to the rewarding of the effort by
the society. The repetition rate is higher among indigenous children than non-indigenous
children (40% versus 31% between ages 8 and 15, and 22% versus 14% between ages 12 and
15).
The difficulty of establishing the sense of causality between aspirations and educational
attainment lies in two potential sources of endogeneity due reverse causality and unobserved
characteristics of the children. In the first case, children may adapt their aspiration to their
performance at school. If they are successful at school, they may target higher educational
attainment and more prestigious occupation. On the contrary, school failure may lead to
downward revision of their aspiration. In the second case, aspiration levels may depends on
unobserved variables that also affect educational attainment. One of these unobserved
variables is ability, as children may adapt their aspiration to what they feel they are able to
reach, and ability is a main factor of the educational performance.
20
To address this difficulty, we rely on two strategies. First, we use the longitudinal dimension
of the data to face the reverse causality problem. We observe the aspiration gap before the
educational performance. We also control by the perception of the parents on the relative
position of their child in her class when the child is 8 years old: “How would you rate your
child’s school performance? Good, average or bad?”, supposing that a large part of the self-
perception of the children on their ability and their past academic progress when they are 8
years old is captured by this variable. To estimate the performance of the children between
age 12 and 15, we control by the score in mathematics at age 12, which is supposed to capture
the school performance of the children before they are 12 years old.
We rely then on instrumental variables estimation to tackle the endogeneity issue due to
unobserved characteristics. We have at our disposal an instrument that could be correlated to
aspirations and independent of school performance. It is the share of the employed labour
force in the district where the child is living, that has an occupation providing a low socio-
economic status. More precisely, the instrument is the share of the employed labour force in
the district that has an occupation which score S belongs to the group 1 ( ) defined in
the section 2. To compute this instrument, we use the 2007 Population Census (Censos
Nacionales 2007 XI de Población y VI de Vivienda), where people are asked about their main
job. For each district where children of Young Lives survey are living, we impute a score to
each worker following the same methodology as in section 2, that allow us to define to which
group of the worker belongs.6 This instrument reflects the concept of aspiration window
defined by Ray and consequently is supposed to shape the aspiration gap. On the other hand,
it is independent from repetition, at repetition rate is supposed to be not correlated with
district characteristics because it well acknowledged that the teacher adapted to the level of
the classroom when evaluating the pupils in such a way as to maintain approximately the
same normal distribution of marks from year to year (De Landsheere 1980). Repetition
reflects then the effort of the children relatively to her classmate, and then it independent from
the context at the district level.
We also partially control for unobserved ability by introducing the score of Raven test at age
8. This test is an intelligence test, not an achievement test, and it does not rely on language to
determine cognitive abilities. We also introduce an anthropometric measurement of the child
6 These groups are the same than for aspiration. It means that they are defined as the quartiles of the score from
all the occupations listed by the children in R1, R2 and R3 when they are asked about their aspirations.
21
at age 8 to control the health status of the child. This variable takes the value 1 if the child is
malnourished (height-for-age z-< -2). The other controls of the models are the same than in
model (2).
The equations we estimate are then:
(4)
(5)
where is the educational outcome of child i at time t, is the instrumental variable.
Models 11 to 13 estimate the probability of repeating at least one year between age 8 and age
15 by the aspiration gap at age 8 and controls defined earlier. Models 14 to 22 estimate the
probability of repeating at least one year between age 12 and age 15 by the aspiration gap at
age 12 and controls defined earlier. Models 11, 12, 14, and 15 rely on a probit. Models 112
and 14 introduce interaction between the aspiration gap and the ethnic dummy to test whether
the effect of the aspiration gap is different for the indigenous children than for the non-
indigenous children. Models 13 and 16 are based on instrumental variables. Table 7 shows the
results of these estimations.
In all the models, the aspiration gap has a significant effect on the probability to repeat a class,
with the exception of the model 14. However, the sense of the effect differs according to the
models. It is negative in the model 11 based on a probit specification without interaction,
meaning that children who have very high aspiration compared to their current social status
have a lower probability to repeat, other things being equal. However, looking at the results of
models 12 and 15 which include interactions, it appears that this negative effect is driven by
the effect of aspiration gap for non-indigenous children. Indeed, aspiration gap has a reverse
effect for indigenous and non-indigenous children. For indigenous children, the highest the
aspiration gap, the highest the probability to repeat a year. However, the positive effect of
aspiration gap on the probability to repeat is significant only between ages 12 and 15 (model
15). It means that indigenous children may not associate an educational level to the
occupation they aspire to in their early age.
Moving to the models 13 and 16, which are best designed to limit the endogeneity bias, the
results shows a positive and significant effect of the aspiration gap on the probability to
repeat. This result is consistent with the ones of the others models as far as we consider the
22
local average treatment effect (LATE). As highlighted by Imbens and Angrist (1994),
instrumental variables gives the effect of the aspiration gap on repetition for those whose
aspiration gap is impacted by the share of workers in the district who have an occupation
providing a low socio-economic status, or in other words, for those who integrate in their
aspiration window people who are on at the bottom of the social ladder. And indigenous
children are more likely to do it than non-indigenous children, as shown by the results of the
regression of the aspiration gap on the instrument (share of workers in the district with low
socio-economic status) and interaction between ethnic dummy and the instrument (not
reported).
To conclude, our estimates suggest that a large aspiration gap at age 12 has a negative impact
on the educational outcomes of indigenous children. Consequently, the highest distance for
indigenous children to fill the gap between the occupation they aspire to and their current
socio-economic status appears to discouraged them from providing effort at school, as the
effort needed to reach this occupation are too costly.
6. Conclusion
In this paper, we examine whether aspirations failure is one of the mechanisms of
intergenerational poverty trap for indigenous people. To address the question, we rely on the
longitudinal Young Lives data, focusing on the older cohort, surveyed at 8, 12 and 15 years
old, in 2002, 2006 and 2009 respectively. We find that aspiration failure is a channel of
inequality persistence between indigenous and non-indigenous people. Indeed, in addition to
highest economics constraints, we show that indigenous children face an aspiration failure
that affects their decision-making process and leads to underinvestment in education.
The aspiration failure do not takes the form of a lack of aspiration. Aspirations of indigenous
children are quite similar to those of non-indigenous children once the level of external
constraints is controlled for. However, indigenous children differ tremendously according to
their aspiration gap, that measure the distance between the socio-economic status they aspire
to reach and their current one. The distance they have to cover to fill this gap is nearly two
times higher than non-indigenous children. But the difference is driven by the highest external
constraints of indigenous children. This result calls for the rejection of the hypothesis that
indigenous children have internalized racial schemas that incorporate beliefs about occupation
23
or about their opportunities, and which would lead to a lack of aspiration for indigenous
children.
In addition, this paper shows that large aspiration gap impede indigenous children to adopt
forward-looking behaviours, such as investment in education. This paper adopts an original
strategy to identify the causal effect of aspiration gap on educational outcomes. It relies on an
instrumental variable calculated using the Population Census, that is the proportion of the
workers with low socio-economic status in the district where the child is living. Our estimates
suggest that the aspiration gap for indigenous children is too large, in so far as it has a positive
effect on the probability to repeat a grade. The disincentive effect of large aspiration gap on
the effort provided at school becomes more important while the indigenous children grow up,
as they may be more aware of the distance between their aspiration and their current status.
Consequently, policy that aims at alleviating the exogenous socio-economic constraints faced
by indigenous people will contribute to bridge their aspiration gap, and therefore will have an
incentive effect on the effort they provide to improve their social-economic status. The
reduction of the aspiration gap then has a multiplier effect on policy which seeks to break the
vicious circle of poverty for indigenous people.
24
References
Acemoglu, D., Johnson, S., & Robinson, J. (2002). Reversal of Fortune: Geography and Institutions in
the Making of the Modern World Income Distribution. The Quarterly Journal of Economics, 117(4),
1231-1294.
Adams, N., & Valdivia, N. (1991). Los nuevos empresarios. Etica de migrantes y formación de
empresas en Lima. Lima: IEP.
Ames, P. (2012). Language, culture and identity in the transition to primary school: challenges to
indigeneous children’s rights to education in Peru. International Journal of Educational Development,
32(3), 454-462.
Ansión, J., Lazarte, A., Matos, S., Rodriguez, J., & Vega-Centeno, P. (1998). Educación : la mejor
herencia. Decisiones educativas y expectativas de los padres de familia. Una aproximación empírica.
Lima: Pontifica Universidad Católica del Perú.
Appadurai, A. (2004). The capacity to aspire. In M. Walton & V. Rao (Eds), Culture and Public
Action (59-84). Stanford, CA: Stanford University Press.
Barrón, M. (2005). Cuánto cuesta ser provinciano a un empleado de Lima Metropolitana: Una
aproximación mediante Propensity Score Matching. Observatorio de la Economia Latinoamericana,
Grupo Eumed.net (Universidad de Malaga), issue 47.
Barrón, M. (2008). Exclusion and Discrimination as Sources of Inter-Ethnic Inequality in Peru.
Economía, 31(61),51-80.
Benavides, M., Olivera, I. & Mena M. (2006). De papás y mamas a hijos e hijas: las aspiraciones
sobre el futuro y rol de las familias en las actividades escolares en el Perú . In M. Benavides (Ed.) Los
desafíos de la escolaridad en el Perú (157-201). Lima: GRADE.
Bernard, T., Dercon S., & Tafesse A.S.0 (2011). Beyond Fatalism-an Empirical Exploration of Self-
efficacy and Aspirations Failure in Ethiopia. CSAE Working Paper 2011-03.
Bernard T., & Tafesse A.S. (2012). Measuring Aspirations: Discussion and Example from Ethiopia.
IFPRI Discussion Paper 01190.
Bigler, R.S., Averhart, C.J., & Liben, L.S. (2003). Race and the Workforce: Occupational Status,
Aspirations, and the Steeotyping Among African American Children. Developmental Psychology,
39(3), 572-580.
Bourdieu, P. (1997). Méditations pascaliennes. Paris: Seuil.
Camfield, L., Masae A., Mc Gregor. A., & Promphaking B. (2012). Culture of Aspiration and
Poverty? Aspirational Inequalities in Northeast and Southern Thailand. DEV International
Development Working Paper 38, University of East Anglia.
Cosby, A.G., & Picou, J.S. (1973). Structural Models and Occupational Aspirations: Black-White
Variations Among Deep-South Adolescents. Journal of Vocational Behavior, 3(1), 1-14.
Cueto, S. (2007). Las evaluaciones nacionales e internacionales de rendimiento escolar en el Perú:
balance y perspectivas. En Grupo de Análisis para el Desarrollo (Ed.). Investigación, Políticas y
Desarrollo en el Perú. (405-455). Lima: GRADE.
25
Cueto, S., Guerrera, G. Leon J., Seguin E., & Munoz I. (2009). Explaining and overcoming
marginalization in education: a focus on ethnic/language minorities in Peru, Background paper
prepared for the Education for All Global Monitoring Report 2010.
De Landsheere, G. (1980). Evaluation continue et examen: Précis de docimologie. Brussels: Editions
Labor.
Dercon, S., & Krishnan, P. (2009). Poverty and the Psychosocial Competencies of Children: Evidence
from the Young Lives Sample in Four Developing Countries. Children, Youth and Environments,
19(2), 138-163.
De Soto, H. (1986). El otro sendero. La revolución informal. Lima: El Barranco.
Duflo, E. (2006). Poor but Rational? In A. V. Banerjee, Bénabou, R., & Mookherjee, D., (Eds),
Understanding Poverty (367-378). New York: Oxford University Press.
Elster, J. (1985), Sour Grapes: studies in the subversion of rationality. Cambridge: Cambridge
University Press.
Engerman, S.L., & Sokoloff, K.L. (2005). Colonialism, Inequality, and Long-run Paths of
Development. NBER Working paper N°11057.
Escobal, J., Lanata, C., Madrid, S., Penny, M., Saavedra, J., Suarez, P., Verastegui, H., Villar, E., &
Huttly, S. (2003). Young Lives Preliminary Country Report: Peru. Oxford, UK: Young Lives.
Escobal, J., & Ponce C. (2007). Economic Opportunities for Indigenous Peoples in Rural and Urban
Peru. In: Conference Edition: Economic Opportunities for Indigenous Peoples in Latin America.
Washington DC: World Bank.
Escobal, J., & Flores, E. (2008). An Assessment of the Young Lives Sampling Approach in Peru.
Young Lives Technical Note n°3.
Figueroa, A. (2006). El problema del empleo en una sociedad sigma. CISEPA Working Paper N° 249.
Lima: Pontificia Universidad Católica del Perú.
Ganzeboom, H.B., & Treiman, D. (1996). Internationally Comparable Measures of Occupational
Status for the 1988 International Standard Classification of Occupations. Social Science Research.
25,201-239.
Glick, J. E., & White, M. J. (2004). Parental Aspirations and Post-Secondary School Participation
among Immigrant and Native Youth in the United States. Social Science Research, 33, 272-299.
Goldenberg, C., Gallimore, R., Rees, L., & Garnier, H. (2001). Cause or effect? A longitudinal study
of immigrant latino parents' aspirations and expectations, and their children's school performance.
American Educational Research Journal, 38(3), 547-582.
Golte, J., & Adams, N. (1987). Los caballos de Troya de los conquistadores. Estrategias campesinas
para la conquista de la gran Lima. Lima: IEP.
Gottfredson, L. S. (1981). Circumscription and compromise: A developmental theory of occupational
aspirations. Journal of Counseling Psychology, 28(6), 545-579.
26
Henríquez, N. (1995). La sociedad diversa, hipótesis y criterios sobre la reproducción social. In G.
Portocarrero & M. Valcarcel (Eds.), El Perú frente al Siglo XXI. Lima: Pontifica Universidad Católica
del Perú Fondo Editorial.
Hoff, K., & Pandey, P. (2006). Discrimination, Social Identity, and Durable Inequalities. American
Economic Review, 96(2), 206-211.
Hubert, L. (1997). Etnicidad y economía en el Perú. IEP, Documento de Trabajo n°83.
Ibrahim, S. (2011) Poverty, Aspirations and Wellbeing: Afraid to aspire and Unable to Reach a better
Life – Voices from Egypt. BWPI Working Paper no. 141, University of Manchester.
Imbens G.W., Angrist, D. (1994). Identification and Estimation of Local Average Treatment Effects.
Econometrica, 62(2), 467-475.
Instituto Nacional de Estadística e Informática (INEI). (2011). Perú: Perfil de la Pobreza por
departamentos, 2001-2010. Lima :INEI.
Jacobs, J.A., Karen, D., & McClelland, K. (1991). The Dynamics of Young Mens Career Aspirations.
Sociological Forum, 6(4), 609-639.
Kao, G., & Tienda, M. (1998). Educational Aspirations of Minority Youth. American Journal of
Education, 106(3), 349-384.
Kuvlesky, W., & Patella, V. (1971). Degree of Ethnicity and Aspirations for Upward Social Mobility
Among Mexican American Youth. Journal of Vocational Behavior, 1(3), 231-244.
Mac Isaac, D. (1994). Peru. In G. Psacharopoulos & H. Patrinos (eds.), Indigenous People and
Poverty in Latin America: An Empirical Analysis (165-204). Washington, D.C.: The World Bank.
MacMahon, M., & Patton, W. (1997). Gender differences in children and adolescents’ perceptions of
influences on their career development. School Counselor, 44, 368-376.
Mahoney, J. (2003). Long-run Development and the Legacy of Colonialism in Spanish America.
American Journal of Sociology, 109(1), 50-106.
Matos Mar, J. (1984). Desborde popular y crisis del Estado. El nuevo rostro del Perú en la década de
1980. Lima: IEP.
Mendoza, O. (1997). Las redes sociales y el crecimiento de las pequeñas empresas en Gamarra, 1980-
1996. In Balbi (Eds.), Aspiraciones, reconocimiento y ciudadanía en los noventa (29-55). Lima:
Pontificia Universidad Católica del Perú.
Mookherjee, D., Napel, S., & Ray, D. (2010). Aspirations, Segregation, and Occupational Choice.
Journal of European Economic Association, 8(1), 139-168.
Ñopo, H., Saavedra, J., & Torero M. (2004). Ethnicity and Earnings in Urban Peru. IZA Discussion
Papers 980. Bonn, Germany: Institute for the Study of Labor.
Oluigbo, N.O. (1976). Patterns of Occupational Choice and Aspiration in Conditions of Economic and
Technological Underdevelopment. Journal of Vocational Behavior, 8(2), 133-44.
Page, L. (2005). From social inequalities to educational inequalities: educational choices and prospect
theory. Revue économique, 56(3), 615-623.
27
Page, L., Levy Garboua, L., & Montmarquette, C. (2007). Aspiration levels and educational choices :
An experimental study. Economics of Education Review, 26(6), 747-757.
Pasquier-Doumer L. (2002). La evolución de la movilidad educacional intergeneracional en el Perú
desde hace un siglo. Bulletin de l’Institut Français d’Etudes Andines, 31(3), 429-471.
Piketty, T. (1998). Self-fulfilling beliefs about social status. Journal of Public Economics, 70(1), 115-
132.
Qian, Z., & Blair, S.L. (1999). Racial/Ethnic Differences in Educational Aspirations of High School
Seniors. Sociological Perspectives, 42(4), 605-625.
Rao, V., & Walton M. (2004) Culture and Public Action. Stanford: Stanford University Press.
Ray, D. (2006). Aspirations, Poverty and Economic Change. In A. Banerjee, R. Bénabou, & D.
Mookherjee (eds), Understanding Poverty. London: Oxford University Press.
Reid, P.T., & Stephens, D. S. (1985). The roots of future occupations in childhood: A review of the
literature on girls and careers. Youth and Society, 16(3), 267–288.
Reissman, L. (1953). Levels of Aspiration and Social Class, American Sociological Review, 18(3),
233-242.
Roemer, J.E. (1998). Equality of Opportunity. Cambridge: Harvard University Press.
St-Hilaire, A. (2002). The Social Adaptation of Children of Mexican Immigrants: Educational
Aspirations Beyond Junior High School. Social Science Quarterly, 83(4), 1026-1043.
Sen, A. K. (2004). ‘How does culture matter?’ in V. Rao and M. Walton (eds.), Culture and Public
Action, Stanford, CA: Stanford University Press.
Trivelli, C. (2005). Una mirada cuantitativa a la situación de pobreza de los hogares indígenas en el
Perú. Economía, 28(55-56), 83-158.
Zhang, Y., Haddad, E., Torres, B., & Chan, C. (2011). The Reciprocal Relationships Among Parents’
Expectations, Adolescents’ Expectations, and Adolescents’ Achievement: A Two-Wave Longitudinal
Analysis of the NELS Data. Journal of Youth and Adolescence, 40(44), 479 - 489.
28
Table 1: Characteristics of the groups of occupational aspirations
# Occu-
pations Mean Std. Dev. Min Max
Low aspirations ( )
Average income (S/.) 14 5000 2197 1964 10286
Average years of education 14 8.0 1.2 5.2 9.4
Standardized score 14 -1.91 0.46 -2.93 -1.40
Intermediate aspirations ( )
Average income (S/.) 14 7150 2382 4155 11798
Average years of education 14 10.9 1.5 8.6 13.4
Standardized score 14 -0.85 0.46 -1.35 0.05
High aspirations ( )
Average income (S/.) 14 12564 2523 8836 16051
Average years of education 14 14.0 1.1 11.9 15.7
Standardized score 14 0.49 0.31 0.06 1.16
Very high aspirations ( )
Average income (S/.) 13 29051 18242 12569 71724
Average years of education 13 16.2 0.5 15.0 16.9
Standardized score 13 2.44 1.31 1.22 5.63
Source: Young lives R1, R2 surveys, ENHAO 2006. Authors’ calculation.
29
Table 2: Distribution of occupations within aspirations groups (%)
At 8 years old At 12 years old Low aspirations (group 1) 12.2 100 Total 4.7 100
Retail trader 1.2 Cook 1.2
Stall saleperson 4 Retail trader 9.5
Agricultural worker 53.5 Agricultural worker 47.3
Fishery worker 0.5 Animal producer 5.3
Handicraft worker in textile 4.5 Taylor 1.8
Backer 0.5 Domestic 26.3
Taylor 0.5 Construction unskilled worker 8.5
Bricklayer 3.7
Street vendor 0.5
Domestic 19.4
Cleaner 4.9
Construction unskilled worker 7
Intermediate aspirations (group 2) 5.4 100 Total 13.7 100
Receptionist 1.1 Decorator 3.3
Sport player 15.4 Sport player 10.3
Secretary 11 Secretary 25
Mechanic 9.4 Waiter 0.4
Carpenter 5.6 Hairdresser 0.4
Construction worker 2.2 Mechanic 40.6
Driver 54.3 Electrician 3.8
Unskilled worker in mining 1.1 Painter 1.7
Carpenter 1.9
Construction worker 0.9
Driver 8.5
Transport worker 3.2
High aspirations (group 3) 51.6 100 Total 43.3 100
Armed forces 0.8 Military officer 0.1
Chief executive 1.4 Armed forces 1.2
Teacher 59.5 Teacher 55.2
Journalist 0.2 Journalist 0.7
Artist 4 Artist 4.4
Religious professional 0.5 Designer 0.4
Designer 0.1 Information technician 0.1
Information technician 0.8 Nursing professional 25.4
Nursing professional 11.9 Accounting associate 0.6
Fire-fighter Police officer 20.7 Guide 0.7
Fire-fighter Police officer 11.2
Very high aspirations (group 4) 30.8 100 Total 38.3 100
Physicist 0.2 Business manager 2.9
Architect 0.2 Physicist 1.1
Engineer 22 Architect 2.5
Biologist 0.2 Engineer 35.1
Medical doctor 70.7 Biologist 1.2
Veterinarian 3.3 Medical doctor 30.3
Accountant 0.2 Veterinarian 3
Lawyer 2.6 University teacher 0.7
Economist 0.3 Accountant 3.7
Aircraft pilot 0.4 Lawyer 16.8
Economist 2.2
Psychologist 0.5
Aircraft pilot 0.2
Total 100 Total 100
N obs 615 N obs 615
Source: Young lives R1, R2. Authors’ calculation.
30
Table 3: Distribution of occupational aspirations by age, sex and ethnic group (%)
At 8 years old
At 12 years old
All I NI G B All I NI G B
Armed forces 0,4 0 0,8 0 0,8 0,6 0 1,1 0 1,1
Business manager 0 0 0 0 0 1,1 0,8 1,4 0,5 1,7
Leader 0,9 1,1 0,8 0 1,9 0 0 0 0 0
Physicist, biologist 0,2 0,1 0,1 0,1 0,1 0,9 0,1 1,6 0,3 1,5
Architect 0,1 0 0,1 0,1 0 0,9 0,6 1,2 1,2 0,7
Engineer 6,8 8,9 4,8 2,9 10 13 14 13 8,7 18
Medical doctor 22 12 31 28 16 12 7,9 15 13 10
Veterinarian 1 0,2 1,8 0,2 1,7 1,1 1,1 1,5 1,9 0,7
Teacher 31 41 21 38 24 24 31 18 28 21
Accountant 0,2 0 0,3 0,3 0 2,5 2,2 2,8 1,6 3,4
Lawyer 0,8 0,9 0,7 0,8 0,8 6,5 6,5 6,4 4,6 8,2
Journalist 0,1 0 0,2 0 0,2 0,3 0 0,6 0,1 0,5
Artist 2,2 3,1 1,3 2,6 1,8 2,6 2,2 2,8 4,2 0,9
Computer operator 0,4 0 0,8 0 0,8 0,1 0 0,1 0,1 0
Aircraft pilot 0,1 0,1 0,1 0 0,2 0,1 0 0,1 0 0,1
Nurse 6,1 7,4 5 11 2 11 15 7,3 21 2,1
Sport player 0,8 0,6 1 0,1 1,5 1,4 0,8 1,9 0 2,7
Secretary 0,6 0,7 0,5 1,2 0 3,4 1,6 5,1 7,1 0
Guide, waiter 0,1 0 0,1 0,1 0 0,4 0,5 0,2 0,8 0
Cook 0 0 0 0 0 0,1 0 0,1 0 0,1
Hairdresser 0 0 0 0 0 0,1 0 0,1 0,1 0
Fire-fighter Policeman 11 7,9 13 3,9 17 4,9 5 4,8 3,3 6,3
Retail trader 0,7 0,8 0,6 1 0,4 1,7 0,3 0,6 0 0,9
Agricultural worker 6,6 5,9 7,1 2,5 10 2,5 1,8 3,2 1,2 3,7
Handicraft worker 0,7 0,5 0,8 1,4 0 0,1 0 0,2 0,2 0
Mechanic 0,5 0,4 0,6 0 1 5,6 5,3 5,8 0 11
Construction worker 0,9 1 0,8 0 1,7 1,1 0,6 1,7 0 2,2
Driver 2,9 2,8 3 0 5,7 1,6 1,5 1,7 0 3,1
Domestic, cleaner 3 3 2,9 6 0,2 0,4 1,2 1,3 1,9 0,6
Labourer 0,9 1,5 0,3 0,1 1,6 0 0,3 0,5 0,2 0,6
Student 0 0 0 0 0 0 0 0 0 0
Total 100
100
100
100
100 100
100
100
100
100
Number of observations
615
237
378
285
330 615
237
378
285
330
Number of occupations 40 26 35 24 28 43 26 42 25 35
Source: Young lives R1, R2 surveys. Authors’ calculation.
Note: I: Indigenous; NI: Non Indigenous; B: Boys; G: Girls. Number of occupations is the number of
different occupations listed by the children and recoded using the 3-digit International Standard
Classification of Occupation 2008 (ISCO-08).
31
Table 4: Distribution of occupational aspirations and average aspiration gap, by age and ethnic group
At 8 years old At 12 years old
I NI Diff I NI Diff
Average aspiration score 0.7 0.8 NS 1.0 1.0 NS
Aspirations groups (%)
Low 12.0 12.2 NS 3.6 5.8 NS
Intermediate 5.2 5.6 NS 10.4 16.7 *
High 60.5 43.3 *** 53.2 34.1 ***
Very high 22.2 38.8 *** 32.8 43.3 **
Total 100 100 100 100
Average aspiration gap 1.5 0.8 *** 1.8 1.0 ***
N obs 237 378 237 378
Source: Young lives R1, R2 surveys. Authors’ calculation.
Note: I: Indigenous; NI: Non Indigenous; Diff column tests the significance of the differences between
the I and NI proportion/mean; NS means that the difference is non-significant, *, **, *** that the
difference is significant at the 10, 5, and 1% levels, respectively.
Table 5: Indicators of aspirations mobility between age 8 and 12 and age 8 by ethnic group
Mobility from age 8 to 12
(%) Score differential Group differential
I NI Diff I NI Diff
Upward mobility 42.4 40 NS 31.6 28.4 NS
Downward mobility 31.5 40 NS 19 29 NS
Immobility 26.1 20 NS 49.4 42.6 NS
Total 100 100 100 100
Source: Young lives R1, R2 surveys. Authors’ calculation.
Note: See table 4.
32
Table 6: Estimation of the scores of aspiration and the groups of aspiration at ages 8 and 12
Source: Young lives R1, R2 and R3 surveys. Authors’ calculation.
Dependent
variables
Score of aspiration
Age 8
Score of aspiration
Age 12
Down-
ward
score
between
ages 8
and 12
Group of aspiration
Age 8
Group of aspiration
Age 12
Downwar
d group
between
ages 8 and
12
Models (1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
Independent
variables OLS OLS OLS OLS Probit
Ordered
probit
Ordered
probit
Ordered
probit
Ordered
probit Probit
Indigenous
child -0.186* 0.050 -0.138 0.106 -0.092 -0.283*** -0.086 -0.141* 0.117 -0.237*
(0.116) (0.119) (0.105) (0.109) (0.113) (0.092) (0.100) (0.091) (0.102) (0.123)
Years of edu
of the 0.038* 0.056*** 0.005 0.045** 0.046*** 0.008
most
educated
parent (0.022) (0.021) (0.020) (0.018) (0.017) (0.021)
Rural -0.361** -0.258* -0.172 -0.297** -0.245* -0.208
(0.166) (0.159) (0.150) (0.129) (0.134) (0.158)
Wealth index 0.750** 0.824** -0.714** 0.688** 1.080*** -1.246***
(0.336) (0.380) (0.316) (0.292) (0.324) (0.338)
Girl 0.507*** -0.057 0.262** 0.356*** -0.017 0.118
(0.105) (0.100) (0.105) (0.093) (0.092) (0.113)
Household
head -0.284** -0.008 -0.042 -0.196* 0.025 -0.104
is a female (0.138) (0.129) (0.134) (0.118) (0.121) (0.148)
Constant 0.953*** 0.033 1.192*** 0.266 -0.075 -0.032
(0.069) (0.262) (0.069) (0.292) (0.232) (0.245)
Constant cut1
-1.420***
-
0.618***
-
1.812***
-
0.916***
(0.082) (0.212) (0.100) (0.258)
Constant cut2
-1.123*** -0.297
-
0.982*** -0.040
(0.074) (0.208) (0.075) (0.248)
Constant cut3
0.192*** 1.111*** 0.004 1.019***
(0.063) (0.208) (0.064) (0.248)
Observations 609 609 609 609 609 609 609 609 609 609
R-squared 0.004 0.101 0.003 0.092
log likelihood -1055.03 -1023.78 -1020.66 -992.24 -391.42 -678.12 -646.15 -681.69 -643.81 -327.85
pseudo-R-
squared . . . . 0.016 0.007 0.054 0.002 0.057 0.029
33
Table 7: Estimation of repetition between ages 8 and 12 and between ages 12 and 15
Dependent variables Repetition between ages 8 and 15 Repetition between ages 12 and 15
Models (11) (12) (13) (14) (15) (16)
Independent variables Probit Probit IV probit Probit Probit IV probit
Aspiration gap age 8 -0.079**
0.458***
(0.031)
(0.133)
Aspiration gap age 12
0.008
0.565***
(0.049)
(0.104)
Indigenous child -0.118 -0.308** -0.137 -0.060 -0.317* -0.177*
(0.127) (0.142) (0.098) (0.139) (0.171) (0.095)
Nind*Aspiration gap -0.154*** -0.041
(0.042) (0.048)
Ind*Aspiration gap 0.052 0.155**
(0.053) (0.070)
Raven score age 8 -0.022*** -0.025*** -0.022*** -0.016* -0.017* -0.008
(0.008) (0.008) (0.008) (0.009) (0.009) (0.008)
Parent's perception of
attainement 0.350*** 0.346*** 0.245** 0.017 0.028 0.115
(0.096) (0.096) (0.113) (0.108) (0.109) (0.077)
Girl -0.125 -0.122 -0.275*** -0.204 -0.199 -0.024
(0.116) (0.117) (0.089) (0.126) (0.127) (0.115)
Years of edu of the -0.061*** -0.053** 0.085* -0.021 -0.011 0.104***
most educated parent (0.023) (0.024) (0.047) (0.026) (0.026) (0.031)
Rural -0.333** -0.328** 0.162 -0.461** -0.452** -0.003
(0.161) (0.162) (0.225) (0.191) (0.190) (0.206)
Household head 0.084 0.091 0.049 0.132 0.134 -0.046
is a female (0.151) (0.151) (0.116) (0.159) (0.161) (0.131)
Wealth index -1.298*** -1.374*** -0.176 -0.791** -0.914** -0.261
(0.344) (0.342) (0.574) (0.362) (0.367) (0.365)
Malnourished 0.071 0.057 -0.074 0.132 0.122 -0.012
(0.138) (0.139) (0.113) (0.148) (0.150) (0.108)
Score in maths at 12
-0.001 -0.001 -0.001
(0.001) (0.001) (0.001)
Constant 0.727** 0.821** -1.134* 0.429 0.377 -1.596***
Second stage equation (0.350) (0.357) (0.644) (0.474) (0.470) (0.521)
Shape of low status
occupation in the
district
-0.586*
-0.514*
Constant
(0.356)
(0.296)
3.417***
3.559***
athrho
(0.482)
(0.422)
-1.303*
-1.417**
ln(sigma)
(0.679)
(0.657)
0.544***
0.457***
(0.088)
(0.092)
Observations 586 586 586 586 586 586
log likelihood -319.920 -315.055 -1468.454 -254.914 -252.065 -1352.304
Aspiration failure: A poverty trap for indigenous children in Peru?
This paper aims to contribute to understand the mechanisms underlying the complex exclusion process of indigenous people in Peru, by analysing the role played by aspirations in the investment in education of indigenous children. Aspirations of indigenous children are quite similar to those of non-indigenous children when comparing children with the same socio-economic status. However, indigenous children differ tremendously according to their aspiration gap, that measure the distance between the socio-economic status they aspire to reach and their current one. The distance they have to cover to fill this gap is nearly two times higher than non-indigenous children. But the difference is driven by the highest economic constraints or more limited access to information of indigenous children. This result calls for the rejection of the hypothesis that indigenous children have internalized racial schemas that incorporate beliefs about occupation or about their opportunities, and which would lead to a lack of aspiration for indigenous children.
In addition, this paper shows that large aspiration gap impede indigenous children to adopt forward-looking behaviours, such as investment in education. This paper adopts an original strategy to identify the causal effect of aspiration gap on educational outcomes. It relies on an instrumental variable calculated using the Population Census. Our estimates suggest that the aspiration gap for indigenous children is too large, in so far as it has a positive effect on the probability to repeat a grade. The disincentive effect of large aspiration gap on the effort provided at school becomes more important while the indigenous children grow up, as they may be more aware of the distance between their aspiration and their current status. The results lead us to conclude that aspiration failure is a channel of inequality persistence between indigenous and non-indigenous people. Indeed, we show that indigenous children face a large aspiration gap that affects their decision-making process and leads to underinvestment in education.
www.younglives.org.uk
About Young Lives
Young Lives is an international study of childhood poverty, involving 12,000 children in 4 countries over 15 years. It is led by a team in the Department of International Development at the University of Oxford in association with research and policy partners in the 4 study countries: Ethiopia, India, Peru and Vietnam.
Through researching different aspects of children’s lives, we seek to improve policies and programmes for children.
Young Lives Partners
Young Lives is coordinated by a small team based at the University of Oxford, led by Professor Jo Boyden.
• EthiopianDevelopmentResearchInstitute,Ethiopia
• PankhurstDevelopmentResearchandConsultingplc
• SavetheChildren(Ethiopiaprogramme)
• CentreforEconomicandSocialSciences,AndhraPradesh,India
• SavetheChildrenIndia
• SriPadmavathiMahilaVisvavidyalayam(Women’sUniversity),AndhraPradesh,India
• GrupodeAnálisisparaelDesarollo(GRADE),Peru
• InstitutodeInvestigaciónNutricional,Peru
• CentreforAnalysisandForecasting,VietnameseAcademyof SocialSciences,Vietnam
• GeneralStatisticsOffice,Vietnam
• Universityof Oxford,UK
Contact:Young LivesOxford Department of International Development,University of Oxford,3 Mansfield Road,Oxford OX1 3TB, UKTel: +44 (0)1865 281751Email: [email protected]: www.younglives.org.uk